The goal of this clinical trial is to determine whether a digital psychological literacy program delivered by an AI-powered virtual agent is non-inferior to traditional teacher-led instruction in improving mental health literacy among adolescents. The study focuses on middle and high school students aged 11 to 18 years attending a school in Shanghai, China, including those with varying levels of emotional well-being but excluding those with diagnosed psychiatric disorders or severe physical illness. The main questions it aims to answer are: Is the virtual agent-delivered psychological literacy program as effective as teacher-led instruction in improving adolescents' mental health literacy? Does the virtual agent intervention lead to comparable or better outcomes in secondary measures such as depression, anxiety, psychological resilience, sleep quality, and digital well-being? Researchers will compare students receiving 6 weekly sessions from a virtual digital agent to students receiving the same curriculum delivered by trained psychology teachers to see if the AI-based approach achieves similar improvements in mental health knowledge, attitudes, and help-seeking behaviors. Participants will: Complete online assessments at baseline, week 3, week 6 and week 18 covering mental health literacy (UMHL-A), depression (PHQ-9), anxiety (GAD-7), psychological resilience (RSCA), and other well-being indicators Attend one 45-60 minute session per week for 6 weeks, either interacting with the virtual agent or participating in a teacher-led class Optionally take part in a brief satisfaction survey and/or a focus group interview after the intervention to share their experiences
Study Type
INTERVENTIONAL
Allocation
RANDOMIZED
Purpose
HEALTH_SERVICES_RESEARCH
Masking
SINGLE
Enrollment
334
This intervention delivers the identical 6-week psychological literacy curriculum as the virtual agent arm, but through standard face-to-face instruction by certified school psychology teachers. Teachers follow a detailed manualized lesson plan covering the same four thematic modules and learning objectives. Sessions include traditional pedagogical techniques such as didactic teaching, group discussion, role-playing exercises, reflective writing, and in-class activities. Each session lasts 45-60 minutes and occurs weekly in the regular classroom setting. Teachers are trained to adhere to the protocol to ensure content equivalence with the virtual agent condition, though minor natural variations in delivery style, pacing, or classroom interaction may occur. This intervention represents the current standard of practice for school-based mental health education in China and serves as the active control condition.
This intervention consists of a 6-week, school-based psychological literacy program delivered by an AI-powered virtual digital agent. The agent uses real-time animation, speech synthesis, and scripted but context-sensitive dialogue to guide students through evidence-based mental health content aligned with China's national mental health education guidelines. The curriculum includes four modules: (1) psychological foundations and self-awareness, (2) stress and coping skills, (3) interpersonal and family communication, and (4) healthy digital habits, sleep, and future planning. Each weekly session (45-60 minutes) is fully digital, interactive, and delivered via classroom computers. The virtual agent maintains consistent content fidelity across all participants. This intervention leverages AI technologies-including natural language understanding and emotion-aware feedback-but does not provide clinical therapy or real-time personalized adaptation beyond pre-programmed branching logic.
Shanghai Nanhui No.2 Middle School
Shanghai, China
Change in Adolescent Mental Health Literacy
Mental health literacy is assessed using the Universal Mental Health Literacy Scale for Adolescents (UMHL-A), a validated 17-item Chinese-language instrument. Part A contains eight items assessing attitudes toward help-seeking and mental health-related stigma, scored on a 5-point Likert scale (range: 8-40). Part B contains nine items assessing general mental health knowledge and knowledge of specific mental illnesses, with each item scored 0 or 1 (range: 0-9). The total score ranges from 8 to 49. Higher scores indicate greater mental health literacy, including more positive attitudes and better knowledge, whereas lower scores indicate poorer mental health literacy. The Chinese version has demonstrated good reliability and validity among adolescents.
Time frame: Baseline, week 3, week 6, week 18
Change in Depressive Symptom Severity
Depressive symptoms were assessed using the Visual Analogue Scale-Depression (VAS-Depression).Depressive symptoms during the preceding week are assessed using a 10-cm visual analogue scale ranging from 0 ("not depressed at all") to 10 ("extremely depressed"). Higher scores indicate greater symptom severity. Scores are categorized as mild (0-3), moderate (4-6), or severe (7-10). The scale demonstrates good concurrent validity with established depression measures, including the Beck Depression Inventory (r = 0.70-0.78), and test-retest reliability exceeding 0.80.
Time frame: baseline, week 3, week 6, week 18
Change in Anxiety Symptom Severity
Anxiety symptoms were assessed using the Visual Analogue Scale-Anxiety. Anxiety symptoms during the preceding week are assessed using a 10-cm visual analogue scale ranging from 0 ("not anxious at all") to 10 ("extremely anxious"). Higher scores indicate greater symptom severity. Scores are categorized as mild (0-3), moderate (4-6), or severe (7-10). Correlations with established measures, including the State-Trait Anxiety Inventory, range from 0.65 to 0.76.
Time frame: Baseline, week 3, week 6, week 18
Change in Sleep Quality
Sleep quality over the past month is assessed using the PSQI (Pittsburgh Sleep Quality Index), a 19-item self-report instrument that evaluates seven dimensions: subjective sleep quality, sleep latency, sleep duration, habitual sleep efficiency, sleep disturbances, use of sleep medication, and daytime dysfunction. Total scores range from 0 to 21, with higher scores indicating poorer sleep quality.
Time frame: Baseline， week 3， week 6， week 18
Change in Risk of Gaming Disorder
Risk of gaming disorder is measured using the GDSS (Gaming Disorder Screening Scale), an 18-item Likert-type scale based on DSM-5 and ICD-11 criteria. It assesses symptoms such as loss of control over gaming, prioritizing gaming over other activities, and continued use despite negative consequences. Total scores range from 18 to 72, with higher scores indicating greater risk.
Time frame: Baseline, week 3, week 6, week18
Change in Self-esteem
Self-esteem was assessed using the Rosenberg Self-Esteem Scale (RSES). The Chinese version of the Rosenberg Self-Esteem Scale is a 10-item unidimensional measure rated on a 4-point Likert scale, with five reverse-scored items. Total scores range from 10 to 40, with higher scores indicating greater self-esteem. Scores are classified as low (10-20), moderate (21-30), or high (31-40). The Chinese version has shown good internal consistency (Cronbach's α = 0.84-0.88) and test-retest reliability of approximately 0.82.
Time frame: Baseline，week 3， week 6 ，week 18
Change in Perceived social support
Perceived social support was assessed using the Adolescent Social Support Scale (ASSS). The Chinese Adolescent Social Support Scale comprises 17 items assessing support from parents, teachers, classmates, close friends, and the school. Items are rated on a 5-point Likert scale (1 = never to 5 = always), yielding a total score of 17-85; higher scores indicate greater perceived social support. Scores are classified as low (17-39), moderate (40-62), or high (63-85). The scale has excellent internal consistency overall (Cronbach's α = 0.96), with subscale α coefficients of 0.85-0.93 and test-retest reliability of 0.84.
Time frame: Baseline, week 3, week 6, week18
Change in Psychological Resilience
Psychological resilience was assessed using the Resilience Scale for Chinese Adolescents (RSCA). The Resilience Scale for Chinese Adolescents contains 27 items covering two broad domains-personal strength and supportive resources-and five underlying factors. Items are rated on a 5-point Likert scale, with 12 reverse-scored items. Total scores range from 27 to 135, with higher scores indicating greater resilience. Scores are classified as low (27-81), moderate (82-105), or high (106-135). The scale has demonstrated good internal consistency (Cronbach's α = 0.83) and split-half reliability (0.80).
Time frame: Baseline, week 3, week 6, week 18
This platform is for informational purposes only and does not constitute medical advice. Always consult a qualified healthcare professional.